Why did my ManyChat bot stop replying to Instagram DMs?

The first thing to check is whether you're still inside Instagram’s 24-hour messaging window. Instagram only allows automated replies within 24 hours of a user’s last interaction unless you’re using an approved message type. If someone messages you after that window closes, the bot will appear silent even though it’s technically running.

If it’s happening on first contact, the next thing to verify is Instagram permissions and account status. Instagram can silently revoke messaging permissions if there’s a policy violation, a business verification change, or if the page was disconnected and reconnected. ManyChat won’t always surface this clearly in the UI.

You’ll want to check three places:

1. In ManyChat → Settings → Instagram, confirm the account shows as fully connected.
2. In Meta Business Manager, verify the Instagram account still has messaging permissions enabled.
3. Check for any policy or quality notifications in Meta — even a soft warning can temporarily block automated replies.

Exactly. That’s one of the most confusing parts. Instagram doesn’t always send explicit error messages back to automation platforms. From your perspective, the bot looks “online,” but messages are quietly blocked at the platform level.

Yes — two other frequent causes are: • Token expiration (especially after password or security changes) • Flow entry conditions that no longer match, for example if a keyword trigger was removed or limited to a specific message type It’s worth sending a test DM from a fresh account to rule out trigger logic issues.

That’s the underlying challenge. Instagram automation is tightly coupled to platform policies, and when messaging rules or permissions shift, bots that rely on direct platform triggers can stop responding without warning. Many teams only find out after customers complain.

Some teams move toward setups where the chatbot layer actively monitors platform delivery failures, falls back to alternative channels, or queues messages until policy conditions are met. Architecturally, that means treating Instagram as one channel — not the entire system — so a single policy change doesn’t silently break customer communication. That’s the kind of approach platforms like SmartCog are designed around, especially for businesses that can’t afford missed messages.

Correct. In most cases, the solution is restoring permissions, confirming policy compliance, and ensuring your automation isn’t fully dependent on a single platform behavior. Once those are in place, replies usually resume without changing the actual conversation logic.